A HOLM starter guide

[Zvuchniak]

That is wrong and could possibly be why your having problems. It is very hard for me to gate that low indoors due to ceiling and floor reflections.
See post #5

Killa... Speed of sound is about 344m/s

1s/344m= 0.0029 which is approximately 3ms per meter (sound travels 1m in 3ms). Math is clear. In the post No.5 it says that there is 4ms first reflection. That tells me that first reflective surface in that apartment was 1.33m (52") away. Math is very clear

[killa]

Killa... Speed of sound is about 344m/s

1s/344m= 0.0029 which is approximately 3ms per meter (sound travels 1m in 3ms). Math is clear. In the post No.5 it says that there is 4ms first reflection. That tells me that first reflective surface in that apartment was 1.33m (52") away. Math is very clear

No need to school me. I know the math but to me your gate looks too low. That dip in the frequency response is more than likely caused by a reflection which would mean the gate is too low. I could tell a lot more if I could get a pic with the window a little less zoomed in on the gate time.

[rpb]

Try changing the scale to 5 dB divisions, and select 200 hz to 20k for the width, and about -5 dB to -70 dB for the height. Try auto zoom on the impulse, or select about a 10 ms section of the impulse window, and then click auto zoom. The 3ms gating should be OK, or close to OK. Make sure that the computer, or another speaker is not also generating the sweep tone.

[killa]

Here's the pic:

Attachment 27178

I'm sorry I did not notice this pic earlier. Looks like you may have some noise going on. If you are using a laptop I have heard operating on battery power helps in some cases. Power supplies can cause noise. I assume there was nothing running nearby such as air conditioning

[rpb]

Looking at you measurement, the sweep appears to marginally be above the noise floor of the room. You are at -45 dB or so. Unless there's something odd with the mic or mic pre-amp, you need to be up around -20 dB or so. Looks like the sweep should be louder.

[Zvuchniak]

I'm sorry I did not notice this pic earlier. Looks like you may have some noise going on. If you are using a laptop I have heard operating on battery power helps in some cases. Power supplies can cause noise. I assume there was nothing running nearby such as air conditioning

No AC is running. I used laptop and desktop and both of them shows the same result

Looking at you measurement, the sweep appears to marginally be above the noise floor of the room. You are at -45 dB or so. Unless there's something odd with the mic or mic pre-amp, you need to be up around -20 dB or so. Looks like the sweep should be louder.

Hi RPB. I tried it but it is the same. Only thing that i could do is to turn up mic gain a bit more but watch out for clipping (red LED lights up when it does). I'll do that and then im off to electronic part store to buy XLR and RCA to make loop measurement cable.

By the way... This is great tutorial you made Thanks

[rpb]

I'm glad you like it! Maybe take a look at the soundcard, or PC level settings on the output / input sensitivities.
If your sweep is relatively loud, then the response on the graph should be higher than it is. My loud sweeps are up
around -10 dB. If I changed the settings in the PC, I could make them show much lower like yours, but that would likely hurt
the signal to noise ratio. So I suspect those are the settings you need to adjust. Soundcards can be tricky to deal
with. The built in special effects sometimes are difficult to turn off. I had several that just were not usable.

[ernperkins]

I've always wondered how to interpret the dB values when looking at the distortion graphs. If the signal is at -10 dB and the distortion is at -50 dB is the actual distortion -40 dB (i.e. in relation to the signal) or -50 dB (i.e. absolute)? Haven't seen this mentioned anywhere.

regards....

[Zvuchniak]

I'm glad you like it! Maybe take a look at the soundcard, or PC level settings on the output / input sensitivities.
If your sweep is relatively loud, then the response on the graph should be higher than it is. My loud sweeps are up
around -10 dB. If I changed the settings in the PC, I could make them show much lower like yours, but that would likely hurt
the signal to noise ratio. So I suspect those are the settings you need to adjust. Soundcards can be tricky to deal
with. The built in special effects sometimes are difficult to turn off. I had several that just were not usable.

I turned it up and reached -20dB without clipping from mic. Here it is:



Frequency looks better but impulse is still bad. I bought connectors and making cable now. I will report later

[ryanbouma]

but impulse is still bad.

Hey, I'm tuxedocivic from AVSforum. I think I might be who sent you over here.

I've had this exact problem before with a laptop. I've never really put my finger on it, but I believe it's the mic in the computer just won't go away and takes the measurement. These mics often have a 5khz low pass, and that looks to be the case here also. They also have a messy impulse response and give that exact output. Same with my webcam on my desktop. Try finding that thing and turn it off. I've become very frustrated with them at times. They just won't go away, even when I disable them it seems.

By the way, you don't select your gate time based on the distance from the mic to the speaker but the delta of distance from first reflection vs direct sound. This should be visible on the IR and there's no need to do the math. The math can be a nice sanity check, but that IR is not right.

Good luck.

[pnoneal...]

Well, I downloaded the 64bit version and it installed to Windows 7 without a problem. I've used SpeakerWorkshop in the past and while old it did install on Windows7 but give an message unable to upadate registry but still seems to function well.

Just curious why they didn't include some code to take impedance measurements and also to measure caps, resistors, and coils and of course crossover modeling?

When I get back from vacation and settled into a new house, I'll check this new software out.

[pnoneal...]

I turned it up and reached -20dB without clipping from mic. Here it is:



Frequency looks better but impulse is still bad. I bought connectors and making cable now. I will report later

It's a tricky process setting up the sound card inputs to proper levels. I use RMAA53 and QuickMixIn to help in this process. Here is a page where you can DL those programs to help set things up properly.
http://www.claudionegro.com/
just go to the downloads page

[Zvuchniak]

Hey, I'm tuxedocivic from AVSforum. I think I might be who sent you over here.

I've had this exact problem before with a laptop. I've never really put my finger on it, but I believe it's the mic in the computer just won't go away and takes the measurement. These mics often have a 5khz low pass, and that looks to be the case here also. They also have a messy impulse response and give that exact output. Same with my webcam on my desktop. Try finding that thing and turn it off. I've become very frustrated with them at times. They just won't go away, even when I disable them it seems.

By the way, you don't select your gate time based on the distance from the mic to the speaker but the delta of distance from first reflection vs direct sound. This should be visible on the IR and there's no need to do the math. The math can be a nice sanity check, but that IR is not right.

Good luck.

Hi Ryan (Tuxedocivic ) Yes, you sent me here

I am measuring with desktop computer. It doesn't have built in mic. The only thing i thought that has to do with this is onboard sound card that i disabled in BIOS and uninstaled drivers so that only thing that i can do settings for is M-audio.

You are right about selecting time gate based on delta of distance from first reflection but only if you can see the reflection. I did it this way because i can't see the first reflection. By setting the time gate at exactly the distance from mic and if i remove reflecting surfaces at at least that same distance i eliminate every posible reflection even if i don't see it.

You were talking about this

[ryanbouma]

so that only thing that i can do settings for is M-audio.

Well in that case, I don't understand why your IR looks so bad. What I see in your IR resembles a poor microphone and soundcard with processing, ime. You're equipment is good.

Is your microphone on a stand? I've seen it before when a microphone has a very nearby reflection. Example. The mic is placed ground plane but isn't directely on the ground. That's about the only other time I've seen an IR look that way.

Sorry

EDIT - can you show a smoothed (no gating) response? And what driver is this? Is the FR known? I still feel this isn't a measurement from you mic.

[Zvuchniak]

It is on the stand. Guys from DIY suggested me to do the loop measurement. To connect the output of the sound card to the input and measure it. It should be flat line but it looks like this:



There is some problem. Big question is can i find it and fix it

[rpb]

I'd suggest looking closely at the settings in the Windows mixer. Mute everything except (I think), the wave files.
Look at input and output level adjustments too. There might be a mic boost as well. Sometimes it's hard to find. I think
I mentioned this in the guide. It looks like maybe you are getting a loop of some kind. Like if the HOLM signal output that goes
to the speaker amp is being fed back into the input, and then being sent back out again to the speaker amp again, and then into
the input again. Each time with a very slight delay. Or it could possibly be a sound card reverb effect of some kind. I
had a soundblaster card that I could not turn the effect off. I currently use the soundcards that came with my very basic
Compac, and Asus computers. One of my computers internal soundcard looked like a brick wall filter at about 7k.

[Zvuchniak]

Since i disabled sound card and uninstaled drivers my mixer looks like this:

[ryanbouma]

Big question is can i find it and fix it

Yeesshh! Ya, it should be a flat line along the 0db axis iirc. Something is screwed up with the settings of the sound card.

Do you have any other sound cards that could power that mic? How about another mini jack mic that you could try using the onboard soundcard.

[Zvuchniak]

With help of NEO Dan and Pano i fixed it. It looks like this now:



But impulse is still screwed up.

[Zvuchniak]

EDIT - can you show a smoothed (no gating) response? And what driver is this? Is the FR known? I still feel this isn't a measurement from you mic.

Sorry now. I didn't see when you edited your post.

Midbas is Eminence Deltalite II 2512 and ungated smoothed response 1/12 looks like this:



Bare in mind that screenshots from above are in 2dB increments. Mic is Ok because i measured B&C DE250 and it was similar to factory, Zilch and BWaslo measurements.